AWS has announced a new offering, AWS AI Factories, designed to deliver dedicated AI infrastructure directly inside customer data centers. The service combines the latest NVIDIA accelerated computing platforms, Trainium chips, AWS AI services, and high-speed networking to enable organizations to develop and deploy AI applications at scale without building their own infrastructure from scratch.
Organizations in regulated industries and the public sector face a critical challenge: deploying large-scale AI projects requires massive investments in GPUs, data centers, and power, along with complex procurement cycles and integration of AI models and services. AWS AI Factories address these hurdles by providing a managed, secure, and low-latency AI environment while leveraging customers’ existing data center space, power, and network connectivity.
This new service operates like a private AWS Region within a customer’s data center, offering managed access to AI infrastructure, storage, databases, and AI services such as Amazon Bedrock and SageMaker. Customers gain immediate access to cutting-edge AI chips, tools, and foundation models without negotiating multiple contracts or navigating long deployment timelines.
AWS and NVIDIA’s collaboration powers the AI Factories. By combining NVIDIA’s Grace Blackwell and Vera Rubin architectures with AWS’s Nitro System, Elastic Fabric Adapter networking, and upcoming NVLink Fusion chip interconnects, customers can accelerate AI workloads with improved performance and efficiency. Ian Buck, NVIDIA’s VP of Hyperscale and HPC, noted that the integrated solution allows organizations to focus on innovation rather than system integration.
The public sector is a key focus. AWS AI Factories meet rigorous security standards for handling workloads across all classification levels, from Unclassified to Top Secret. A first-of-its-kind deployment is underway in Saudi Arabia through a partnership with HUMAIN, featuring an AI Zone with up to 150,000 AI chips and full AWS-managed AI infrastructure. HUMAIN CEO Tareq Amin emphasized that the project will serve both local and global AI demand, creating a scalable ecosystem for AI deployment.
AWS AI Factories exemplify how enterprises and governments can access high-performance AI capabilities faster, more securely, and with lower operational complexity, helping accelerate AI adoption at scale (and pushing their stock up). This comes as companies like Amazon continue to expand their AI investments, including large-scale AI supercomputing projects and internal AI rollouts that have raised concerns about societal impacts
